A Translog stochastic cost frontier has been estimated simultaneously with the economic inefficiency effect model using maximum likelihood for different rice crops. To estimate farm-specific economic efficiencies, single estimation of Translog cost frontier has also been done. Different cost components and their interactions are found to have different impacts on the stochastic cost frontier for different rice crops. The study reveals that there are significant economic inefficiency effect in the production of all rice crops and extension contact has negative impact on the economic inefficiency effect for all rice crops whereas experience has negative impact on the economic inefficiency effect for Boro and Aman rice. For Aus rice, education has positive impact on the economic inefficiency effect. The mean economic efficiencies estimated from Translog cost frontiers for Boro, Aus, and Aman rice are respectively 80%, 60% and 74%. The study also reveals that without change of output the production cost of Boro, Aus and Ainan can be reduced by 20%, 40% and 26%, respectively.